686 INFECTION BY THE BLOOD PROTOZOA 



It is essential in all of this work to keep the original passage strain of trypano- 

 somes continuously in such an animal as the mouse where no crises occur and where 

 the parasites do not come in contact with lytic antibodies and hence do not become 

 biologically changed as they do after each trypanolytic crisis in such hosts as the 

 guinea pig, dog, rabbit, etc. Failure to do this probably explains the peculiar results 

 of Kligler and Weitzman (1924) who failed to find any evidence of the formation of 

 humoral antibodies in experimental trypanosomiasis. They have, however, added 

 the interesting experimental finding that the injection of olive oil brings about re- 

 lapses. In view of the numerous experiments I have already cited, it would seem that 

 their contention, that the defense of the body against trypanosome infections lies 

 in cellular responses rather than humoral antibodies, might be modified simply to 

 indicate that humoral antibodies owe their origin to the cellular responses. In this 

 connection it is interesting to note that Muttermilch (191 1) believes that trypanolysins 

 are formed in the spleen and other hematopoietic organs, and Regendanz and 

 Kikuth (1927) believe that the reproduction-inhibiting reaction product in T. lewisi 

 which is described in the next section is formed chiefly by the reticulo-endothelium 

 of the spleen. 



The acquisition and inheritance of antibody resistance by the trypanosomes is a 

 biological problem of great interest. Similar resistances to antibodies and lethal drugs 

 have been observed in many parasitic and free-living protozoa. (For a general review 

 of these "induced variations" in the trypanosomes see Dobell [191 2], and for similar 

 work on a free-living species, Paramecium, see JoUos [1921].) A remarkable feature 

 of the resistance is that it may be inherited for many asexual generations although 

 in time the strain loses its resistance. Furthermore, a radical change in environment, 

 such as passage through the invertebrate host or through another species of ^•erte- 

 brate host, is followed by the loss of resistance. Trypanosomes, to become resistant, 

 do not have to be in the body, as a few minutes' contact between the organisms and 

 the immune serum produces a resistant strain (Ehrlich, Roehl, and Gulbransen [1909] 

 and Levaditi and Mcintosh [1910]). Among the explanations offered for the mech- 

 anism of this acquisition, the following may be noted: Ehrlich's explanation formed 

 an integral part of his side-chain theory of immunity and is too well known to need 

 discussion. Levaditi and Mcintosh (1910) beheve that it is a process of selection in 

 which the non-resistant strains are killed by the antibody, leaving the resistant ones. 

 If they mean by this the selection of certain genetical strains from a "wild popula- 

 tion" of mixed strains, it is not tenable, as the phenomenon is produced within a single 

 cell strain the individuals of which are presumably of the same genetical constitution. 

 Nevertheless, at each crisis when the resistance is acquired, there is obviously a 

 selection, but one selection is effective — a fact in marked contrast to the slow effects 

 of selection observed by various authors working with the free-living protozoa (see 

 Jennings, 1920). Finally, Rosenthal (1913) believes that there is a substance in the 

 serum which has the specific property of producing relapse strains and which can be 

 dift"erentiated from the lytic antibody by (i) the lack of parallelism between the titre 

 of the two, (2) differences in thermolability, and (3) differences in the fractions of 

 serum in which they are precipitated. 



It would seem probable from the work just presented that if lytic serum were 



